A. Project Summary Acute myeloid leukemia (AML) is an aggressive and often fatal hematologic neoplasm. Despite cytotoxic chemotherapy and allogeneic transplantation, patients often succumb to resistant disease, which is reflected by a poor 5-year survival rate of between 30-40%. Numerous studies suggest that AML is organized in a cellular hierarchy with a rare, self-renewing, leukemic stem cell (LSC) giving rise to the remaining leukemic blasts. The AML LSC is thought to initiate and maintain leukemia in humans, and therefore needs to be eliminated in order to achieve cure. Unfortunately, LSCs are poorly defined, which has limited the ability to directly connect LSCs with poor clinical outcomes. It is not known if LSCs are prevalent in the majority of AML patients, and if they express specific surface markers that can be used for disease monitoring and/or as therapeutic targets. There is a critical need to improve our methods for isolating and studying these clinically relevant cells in AML. The long-term goal is to improve AML outcomes by developing cell-specific monitoring strategies and patient-specific anti-leukemia therapies. To address this goal, the next step and overall objective of this application is to rigorously dissect the cellular heterogeneity of AML through single cell gene and protein expression analysis and large-scale in silico cytometry. The central hypothesis is that LSCs are enriched in rare, disease maintaining, AML sub-populations and can be purified using cell specific surface markers. This hypothesis was formulated based on single cell analysis of primary AML patient samples. A rare sub-population of AML cells was identified that is enriched for stem cell gene expression programs and whose gene expression signature is correlated with shorter overall survival. The rationale for the proposed research is that identifying and purifying LSCs are necessary for understanding leukemogenesis and treatment resistance, and in turn designing curative therapies. Guided by strong preliminary analysis, the central hypothesis will be tested by pursuing two specific aims. The first aim will rigorously dissect the cellular heterogeneity of AML using combined single cell transcriptomic and proteomic analysis in order to design a surface marker panel to isolate and study AML LSCs. The second aim will construct a comprehensive cellular landscape of human AML using large scale cellular deconvolution of bulk gene expression data. The overall contribution is expected to be the development of an actionable LSC immunophenotyping panel, and discovery of a newly defined, clinically relevant LSC-enriched sub-population in human AML. This contribution will be significant because it will allow us to not only improve patient outcomes by optimizing disease monitoring strategies but it will also facilitate the design of curative, patient-specific therapies in AML.